A carbon monoxide alarm sounds, yet the site has a hydrogen-handling process rather than combustion equipment. Someone suggests continuing work because it is “a false alarm caused by hydrogen.” However, possible interference in a reading and a conclusion that the site is safe are separate matters.

Portable gas detector cross-sensitivity means a sensor’s response to substances other than its target gas changes the display. Consider both positive interference, which raises readings, and negative interference, which lowers them. When an alarm occurs, follow prescribed exit and reporting procedures. Once safety is secured, compare the actual sensor part number, official interference table and process substances to investigate causes.

Cross-sensitivity is not gas identification

Electrochemical sensors estimate concentration using target-gas reactions, but their selectivity is not absolute. Other gases can cause a target-gas channel to respond. That does not mean its displayed value can be read as the interfering gas’s actual concentration.

Honeywell’s RAE sensor technical note TN-114 explains positive and negative cross-sensitivity. It also states production-batch differences and limitations for gas mixtures. Read table values as conditional data for the specified sensor, not a universal formula for calculating the composition of site mixtures backward. Honeywell TN-114

Simultaneous responses on several channels do not establish gas identities either. Several gases may actually coexist, or interference, environmental changes and sampling problems may overlap. Verify cause hypotheses by connecting measurement results with process information.

Positive and negative interference create different errors

Positive interference raises a channel’s response even when its target gas is absent or present only at a low concentration. Negative interference lowers target-gas response. The latter can prevent an absence of alarms from supporting an absence of danger.

Phenomenon

Interpretation error

Additional verification needed

Another gas raises the display

Concluding it is the target-gas concentration

Suitable verification methods for target and interfering gases

Another gas lowers the display

Using a low reading as evidence of safety

Potential negatively interfering substances and sensor suitability

Several channels respond together

Concluding there are as many gas types as responding channels

Process substances, sequence of events and independent measurement

A gas is absent from the table

Interpreting this as no interference

Ask whether it was unevaluated or data is unavailable

Even a table entry of “0” requires checking test conditions and detection levels. Blank cells may mean no data, which differs from zero influence. A sensor with “reduced interference” does not guarantee zero interference for every gas, concentration and exposure duration.

For hydrogen and CO, check part numbers before sensor names

Possible CO-channel interference in hydrogen-containing environments is a common review question. Do not assume every CO sensor responds to the same degree. Standard, compensated, filtered and different-generation sensors may need different supporting data.

The official MSA ALTAIR 5X PID parts list distinguishes sensor configurations such as standard CO sensors and hydrogen-resistant combinations. This establishes the need to identify actual installed parts; the list alone does not determine hydrogen interference rates or suitability. Confirm performance separately with manufacturer data for that part. MSA sensor and accessory list

Check in order: instrument model, serial number, sensor part number, replacement history and the revision of the reference material in use. Do not copy another instrument’s interference table simply because both housings display CO. Actual hydrogen risk assessment requires a separate plan and equipment suitable for the hydrogen measurement objective.

Turn interference tables into a site review form

Before reading a table, identify which rows and columns represent target gases and interfering gases. Distinguish test-gas concentrations from channel-response units, and read whether percentages refer to the target range or test-gas concentration. Do not convert symbol-only tables into invented numbers.

This form can also support manufacturer inquiries. Leave unverified values marked “verification required.” Record publication and review dates to identify continued use of outdated tables after sensor changes. Do not simply insert expected site-mixture concentrations into tables to change alarm settings.

Review field

Information to enter

Measurement objective

Target gas, work stage, personal monitoring or point measurement

Instrument identification

Model, serial number, sensor part number and manual revision

Coexisting substances

Process inputs, by-products, cleaners, supply gases and supporting material

Manufacturer interference data

Table title, relevant row, test concentration, response direction and units

Unclear applicability conditions

Mixtures, exposure duration, temperature and humidity, filter condition and unlisted substances

Verification plan

Manufacturer inquiry, suitable-sensor review, independent measurement method and owner

Operating decision

Permitted scope for existing instruments, additional equipment and procedure-change approval

NIOSH guidance on direct-reading monitor evaluation also treats interference as a performance evaluation factor. Explaining site suitability requires going beyond the presence of a sensor to describe expected substances and environmental conditions. NIOSH monitor evaluation report

Hypothetical example: alarm review after a cleaning-process change

This is a hypothetical situation, not an actual company case. After a cleaner is changed, detector channels respond differently than usual during some tasks. The responsible person first follows prescribed alarm response, then obtains the new cleaner’s safety data sheet, use times, ventilation status and instrument numbers.

The team checks the sensor table but cannot find the new substance. Instead of recording “no interference,” it sends the sensor part number and substance information to the manufacturer. It also considers appropriate additional measurement methods for evaluating the workplace and does not restart using existing readings alone before the cause is established.

The practical outcome is a procedure to reassess sensor suitability after process changes, rather than settings that reduce alarms. Changes to raw materials, protective gases, process temperatures or waste treatment can also warrant updating expected substance lists. This does not mean every change causes interference, but retaining existing conclusions without review is unsupported too.

Interference corrections to avoid at the site

Do not mentally subtract “the hydrogen portion” based on a table and ignore the alarm. Mixture responses are not guaranteed to add simply, and actual target gas may coexist. Suspected interference is a reason to investigate, not to skip alarm response.

Do not zero in a hazardous area or remove sensor filters to eliminate interference. Calibration procedures and gas sequences also need review for multi-sensor combinations, so prioritize official model procedures and manufacturer confirmation over generic internet descriptions. If replacement is needed, define compatibility and verification scope using post-replacement return-to-service verification.

This article recommends neither model-specific interference rates nor alarm settings. Where part identification and data scope cannot be established, do not deploy equipment on the expectation that interference will be small; prepare a verifiable measurement plan.

Action checklist

  • [ ] Follow site exit and reporting procedures first when alarms occur.

  • [ ] List target gases separately from expected interfering gases.

  • [ ] Verify installed sensor part numbers and replacement histories.

  • [ ] Read the manufacturer table’s rows, columns, units and test conditions.

  • [ ] Review both positive and negative interference.

  • [ ] Do not treat blanks and unlisted substances as “no interference.”

  • [ ] Avoid arbitrary subtraction, zero adjustment or alarm-setting changes based on tables.

  • [ ] Document manufacturer inquiries and additional measurement plans for mixtures and data gaps.

Conclusion

Portable gas detector cross-sensitivity involves both excessive alarms and under-reading hazards. Interference tables have limited interpretation only when sensor part numbers and test conditions match; they are not automatic correction tools for site gas mixtures.

When a reading’s meaning is uncertain, apply safety procedures before establishing causes. Reading interpretation of a reading of 0 with electrochemical sensor operating conditions helps distinguish low numbers from valid normal measurements.

Sources